Neural network for enhancement of end milling processes through accurate prediction of temperature in the cutting zone

Understanding the complex influence of temperature in the cutting zone in order to achieve a successful and efficient machining process is of critical importance. The temperatures generated during processing have a crucial influence on achieving optimal results. In order to achieve the desired result, and to optimize the process, understanding of this complex interaction is necessary. The focus of this research was on the analysis of the influence of temperature in the cutting zone during the milling process. Particular attention has been paid to the influence of cutting speed, feed rate, and cutting depth on temperature generation. The design of the experiment was carried out according to the central composite design, which included 24 distinct cases. The experiment was performed on a Maho 600C vertical milling machine, using 30CrNiMo8 material. Temperature was measured using a FLIR InfraCAM Western thermal imaging camera. Based on the collected experimental data, artificial neural network and mathematical models which define the dependence of temperature on cutting parameters have been developed. The best performance was demonstrated by the artificial neural network with a 3 × 10x5 × 2x1 architecture, trained using the SCG learning algorithm, which resulted in a mean squared error of 0.000782. A comparative analysis was performed between the values gained by experiments and values provided by neural networks and mathematical models. The results showed that the artificial neural network achieved the smallest deviations. These results demonstrate the quality and reliability of our temperature prediction models for machining processes, which allow us to further optimize these processes and achieve the desired results.